Shenzhenaudio Topping EX5 Review (DAC and Headphone Amplifier)

[respice finem]

@Francis Vaughan That's right, however, in a studio, an analog-active crossover has the advantage of no latency (like in the Neumann KH 310A). For home use, if you feed analog signal, it doesn't have to be ADC converted for the crossover and back DAC for the power amps (rather a "cosmetic" issue but still).

 

[JohnYang1997]

@Francis Vaughan That's right, however, in a studio, an analog-active crossover has the advantage of no latency (like in the Neumann KH 310A). For home use, if you feed analog signal, it doesn't have to be ADC converted for the crossover and back DAC for the power amps (rather a "cosmetic" issue but still).

Fixed active crossover is fine. It can be done with extremely low noise and distortion if one wants to. Not the case with PEQ or more complex filters.

 

[Chrise36]

There is exactly no intrinsic difference. Mathematically you can create any analog PEQ section in digital, and within the capabilities of the DAC to resolve, the results will be perfect. Further, there are things you can do in the digital domain that cannot be done in the analog. Maybe the necessary imperfection in the analog implementation create systems some people like. Done right both analog and digital can achieve essentially transparent results. Done wrong and either can be a problem. Cheap digital can be a problem, just as cheap analog can be. In the modern world digital is nearly free, whereas analog is always going to be significant effort. Modern processing power avoids the compromises that may have beset older digital systems. If we demand SOTA, digital wipes the floor with analog. The one place where analog is worth considering is when latency is an issue. For domestic systems this is rarely a problem.

That is all well but in practice engaging DSP on pc (ROON) or EAPO gives me audible degradation it just sounds compressed.I should try hardware solution to compare.

 

[Chrise36]

Fixed active crossover is fine. It can be done with extremely low noise and distortion if one wants to. Not the case with PEQ or more complex filters.

That should suit well. A gain swith for every channel is a must.Please try to develop something like this!

 

[respice finem]

Possibly your overall SPL is lower because the equalizer adjusts it down automatically, if you make "+" corrections.
The slightly louder sound without EQ is then perceived as better.

 

[Chrise36]

Possibly your overall SPL is lower because the equalizer adjusts it down automatically, if you make "+" corrections.
The slightly louder sound without EQ is then perceived as better.

I adjusted the overall gain to the same spl but still...It is less of a problem when adjusting the bass and leave the rest of the frequencies untouched.

 

[gonzoucab]

Analogue PEQ is very noisy.

and active crossovers?

 

[BeerBear]

I think DSP is a whole different area to get into. And then there will be conspiracy theories that the DAC is not bit perfect. "They must be doing something to make their DAC get better reviews." Just look at the MQA debate. IMO, unless there is a built-in headphone amp, PEQ is better done with playback software instead of the DAC.

Software DSP isn't always an option. I can't use it, for example, because I use ASIO [for low latency]. It's a niche use case, though, I'll give you that.

 

[Robbo99999]

That is all well but in practice engaging DSP on pc (ROON) or EAPO gives me audible degradation it just sounds compressed.I should try hardware solution to compare.

Nonsense re DSP via EqualiserAPO or ROON giving audible degredation. More likely to be frequency response related with regard to any changes you're making to the frequency response - it's not gonna be something inherent in the DSP of EqualiserAPO or ROON that would cause, absolutely not.

 

[maxxevv]

I admit not getting lost into property labyrinths in the minotaur's island is a hard task, but I am sure it is not only me who got confused by the misteriously disappeared www.smsl-topping.com.

Topping and SMSL are two completely disparate companies.

That was just an opportunistic seller leveraging on the popularity of the 2 brands.

It had nothing to do with either companies other than selling their products. I pointed out on it when it was raised back then.

 

[Chrise36]

Nonsense re DSP via EqualiserAPO or ROON giving audible degredation. More likely to be frequency response related with regard to any changes you're making to the frequency response - it's not gonna be something inherent in the DSP of EqualiserAPO or ROON that would cause, absolutely not.

I think it is about compressing the dynamics which is more audible to me. I tried many times sometimes it works but most of the time the sound is just more natural without DSP. i found ROON convolution filters horrible to be honest. I could be making something wrong or i just do not like the result and that is not nonsense...That belongs to another thread anyway.

 

[Robbo99999]

I think it is about compressing the dynamics which is more audible to me. I tried many times sometimes it works but most of the time the sound is just more natural without DSP. i found ROON convolution filters horrible to be honest. I could be making something wrong or i just do not like the result and that is not nonsense...That belongs to another thread anyway.

Have a look in the following thread(https://www.audiosciencereview.com/...truth-about-dsp-volume-control-in-roon.23552/), some folks have already measured and proved that software manipulation of volume (and by some extension EQ) will not inherently reduce sound quality. Whatever you're experiencing re degredation of sound quality is gonna be related the frequency response changes you're making (which of course has an audible impact, and by definition in fact!) - ie you're making the wrong frequency response adjustments.

If you're operating your system in 24bit or above then software volume control will not realistically reduce your dynamic range by any kind of noticeable level. Just make sure you're in at least 24bit mode, even if the audio you listen to is 16bit. There is no detriment to running your system at 24bit & above when you listen 16bit recorded music, it doesn't matter that they don't match.......just that running a greater bit depth for your system (ie 24bit or 32bit) will mean you won't lose any appreciable dynamic range when manipulating the volume digitally (which is what you do as part of the process of digital EQ). So make sure your operating system is running in 24bit or 32bit audio mode.

 

[Chrise36]

Have a look in the following thread(https://www.audiosciencereview.com/...truth-about-dsp-volume-control-in-roon.23552/), some folks have already measured and proved that software manipulation of volume (and by some extension EQ) will not inherently reduce sound quality. Whatever you're experiencing re degredation of sound quality is gonna be related the frequency response changes you're making (which of course has an audible impact, and by definition in fact!) - ie you're making the wrong frequency response adjustments.

If you're operating your system in 24bit or above then software volume control will not realistically reduce your dynamic range by any kind of noticeable level. Just make sure you're in at least 24bit mode, even if the audio you listen to is 16bit. There is no detriment to running your system at 24bit & above when you listen 16bit recorded music, it doesn't matter that they don't match.......just that running a greater bit depth for your system (ie 24bit or 32bit) will mean you won't lose any appreciable dynamic range when manipulating the volume digitally (which is what you do as part of the process of digital EQ). So make sure your operating system is running in 24bit or 32bit audio mode.

I have read that all. Can you please enlighten me how DAC IMD looks when we are dealing with lower level signals eg -30 dbfs versus an analog volume control? My problem is not frequency response related as i said. The point of active crossover is eliminating the passive components and using appropriate amps for every driver eg class a or ab for tweeter and clasd d for bass and mid.With gain switch for every driver you can sort of do parametric eq in the analog domain.

 

[Francis Vaughan]

can sort of do parametric eq in the analog domain.

Really by definition this cannot be parametric EQ. That is understood to minimally include variable frequency and variable Q for each section. You get neither. Messing with the gain of individual drivers is never going to substitute for any other EQ, tilt is impossible for a start.
You will never usefully EQ the system unless by improbable blind luck. A lottery win is more likely.
An active crossover is about a synergy of effects. There is much wringing of hands over passive crossover losses and possible distortion. But the gains going active are less than you might hope. Professional sound systems gain with power handling and this can help anywhere. The ability to create more advanced crossover sections is another win.

 

[JohnYang1997]

I have read that all. Can you please enlighten me how DAC IMD looks when we are dealing with lower level signals eg -30 dbfs versus an analog volume control? My problem is not frequency response related as i said. The point of active crossover is eliminating the passive components and using appropriate amps for every driver eg class a or ab for tweeter and clasd d for bass and mid.With gain switch for every driver you can sort of do parametric eq in the analog domain.

If there's no imd hump, -30dB imd will be basically perfect.

 

[Robbo99999]

I have read that all. Can you please enlighten me how DAC IMD looks when we are dealing with lower level signals eg -30 dbfs versus an analog volume control? My problem is not frequency response related as i said. The point of active crossover is eliminating the passive components and using appropriate amps for every driver eg class a or ab for tweeter and clasd d for bass and mid.With gain switch for every driver you can sort of do parametric eq in the analog domain.

If you want to compare IMD of DACS at digitally lowered volume vs that of analog volume control then obviously you're gonna have to look at the specific DAC & amplifier graphs that are displayed on this site for the equipment you wish to compare at the levels that you mention - I'm not gonna do that for you. Analog volume control when done right can retain more SINAD than digital volume control but you'll have to be a judge of whether that is noticeable. In most cases it is a non-issue, and when it is an issue with certain very sensitive headphones for instance it's normally an issue of hearing audible hiss when the noise floor is raised.....hence if you don't hear hiss then it's not a problem. (EqualiserAPO & Roon is not your problem)

EDIT: JohnYang put your mind at ease with specific knowledge of this DAC/amp.

 

[Chrise36]

Really by definition this cannot be parametric EQ. That is understood to minimally include variable frequency and variable Q for each section. You get neither. Messing with the gain of individual drivers is never going to substitute for any other EQ, tilt is impossible for a start.
You will never usefully EQ the system unless by improbable blind luck. A lottery win is more likely.
An active crossover is about a synergy of effects. There is much wringing of hands over passive crossover losses and possible distortion. But the gains going active are less than you might hope. Professional sound systems gain with power handling and this can help anywhere. The ability to create more advanced crossover sections is another win.

Of course it is not the same the
That is why i said sort of maybe band eq is a more appropriate expression. If you have nulls the end result is the same you can not deal with these with DSP also.

 

[Chrise36]

If you want to compare IMD of DACS at digitally lowered volume vs that of analog volume control then obviously you're gonna have to look at the specific DAC & amplifier graphs that are displayed on this site for the equipment you wish to compare at the levels that you mention - I'm not gonna do that for you. Analog volume control when done right can retain more SINAD than digital volume control but you'll have to be a judge of whether that is noticeable. In most cases it is a non-issue, and when it is an issue with certain very sensitive headphones for instance it's normally an issue of hearing audible hiss when the noise floor is raised.....hence if you don't hear hiss then it's not a problem. (EqualiserAPO & Roon is not your problem)

EDIT: JohnYang put your mind at ease with specific knowledge of this DAC/amp.

Thanks for the confirmation.In this speficic dac/ amp IMD will be -90dbfs if i understand right. So it does make sense to add a good implemented analog volume control theoretically at least.

 

[JohnYang1997]

Thanks for the confirmation.In this speficic dac/ amp IMD will be -90dbfs if i understand right. So it does make sense to add a good implemented analog volume control theoretically at least.

Emmm. It depends on what you want to achieve. Take a look at d30pro. It's far better. But I don't think for most cases EX5 will cause audible noise.

 

[Francis Vaughan]

Of course it is not the same the
That is why i said sort of maybe band eq is a more appropriate expression. If you have nulls the end result is the same you can not deal with these with DSP also.

It really isn't clear what you are trying to achieve. Nobody has suggested anything about room nulls. Everyone knows they can't be fixed with EQ, so they are never part of the conversation.

A crossover's job is as the name says - to crossover between drivers. The slopes, turnover points, and special tweaking of sections are designed to work with the specific drivers and cabinet design. Changing the gain on one driver throws a heap of careful design out the window. Any good speaker will have some hidden optimisation in the crossover design, optimisations that will not be obvious. And correct operation depends upon the exact relative output of the different drivers. No crossover is brick-wall, and the out of band response must, and is, considered when considering the overall response of the speaker. In the extreme, if you have a second order crossover all hell breaks loose if the gain is changed.

There seems to a set of rather dated idea and some very specific poor experience with DSP here that is driving a desire for a very specific and unusual (to say the least) approach to system design. It isn't as if this stuff is new or controversial. It is extraordinarily unlikely that varying from the default case of DSP all the way to DACs for with digital volume will not yield the best possible final result. Given a competent modern DAC of almost any price, any noise or distortion products will always be below audibility. Strange hybrid systems will inevitably just yield either a poorer result, or a result that costs more to implement to get the same performance. It was not always so, but the last few years have swung the balance firmly this way.

This isn't to say that it isn't easy to mess it up. We assume competent design. Perhaps, given the history of high-end audio this is not as safe an assumption as one might wish.